4.1.1 Instructional activities related to students’ experiences of situ- ational epistemic emotions
The question concerning students’ experiences of epistemic emotions in different instructional activities was addressed in Study I in particular, but it was also touched upon in Study III. According to the multilevel regression analysis re- ported in Study I, the level of epistemic emotions varies significantly during sci- ence lessons with regard to different instructional activities.
First, orienting and engaging activities, as conceptualised in Study I, appear to be particularly relevant in terms of inducing positive epistemic emotions such as excitement and curiosity, and preventing boredom. According to another study, these activities are also related to aroused interest in students (Vilhunen, Lavonen, et al., 2022). Also, previous research has shown that introductory or demonstration activities can arouse positive emotions in students (King et al., 2015).
In a way, the computer-based learning activity described in Study III could also be considered an engaging and orienting activity: it was implemented as an introductory exercise in the study period, the aim being to engage students in ac- tively observing and coming up with scientific questions. In this regard, the find- ings from Study III corroborate those from Study I: the majority of the students
belonged to the latent class characterised by high levels of curiosity in addition to increasing levels of surprise and confusion.
The above findings can be considered expected. Excitement and curiosity are likely to occur when the aim of the teacher is to orient students towards and engage them in learning. Previous studies have also reported a positive relation between curiosity and engagement (Bosch & D’Mello, 2017), as well as curiosity and knowledge-exploration behaviour (Gruber & Ranganath, 2019; Liquin & Lom- brozo, 2020; Vogl et al., 2020).
Second, cognitive load and taking an active role in learning seem to affect stu- dents’ epistemic emotions. The instructional activities categorised in Study I as
“Conducting investigations and analysing data” and “Developing models” could be considered scientific practices, as described in previous literature (Krajcik &
Shin, 2014; National Research Council, 2012). Together, they could even be clas- sified as “doing laboratory work”. However, in terms of the epistemic emotions they induce, they differ significantly. The cognitively less demanding practical part of laboratory work (conducting investigations) induces relatively low levels of anxiety, confusion and frustration, whereas the computational and conceptual aspects (analysing data and developing models) with a higher cognitive load in- duce significantly more anxiety, confusion and frustration among students.
The role of cognitive load on students’ epistemic emotions seems plausible also based on the results of Study III. The cognitive load of the POE task increases towards the end of the activity as many students experience the information in the last situation as incongruent. Simultaneously, the experienced levels of surprise and confusion among the students seemed to increase. These findings are in line with previous research results indicating an interrelation between cognitive load and negative, activating emotions (Cheng et al., 2020; M. Han & Gutierez, 2021;
Hong et al., 2017) in the context of science education.
Taking an active role in a learning activity could also increase students’ cog- nitive load. Thus, a similar difference as between the practical and computational parts of laboratory work is evident among the more traditional categories of in- structional activity used in Study I, namely “Teacher talk” and “Tasks for skills and content”. The most prevalent epistemic emotion students experience during teacher talk is boredom, which is in line with findings reported in previous re- search (Mann & Robinson, 2009). By contrast, when students work with tasks for skills and content, they experience relatively high levels of anxiety, confusion, and frustration. These activities may both include different levels of cognitive de- mand, from low to high. However, a student’s role in these activities differs sig- nificantly. Students may choose to play a passive role during teacher talk, even when it is interactive, and thus eventually become bored. Tasks, on the other hand, require students to be active learners, which again may increase the cognitive load and further induce negatively valenced epistemic emotions.
Thus, analysing data and developing models, in addition to working on tasks focused on skills and content, can be seen as the most demanding classroom ac- tivities from the students’ perspective: they relate to higher levels of confusion, frustration and anxiety than some of the other instructional activities. Indeed, it was also found in another study that activities requiring computational thinking and active participation are among the most challenging for students (Vilhunen, Lavonen, et al., 2022).
The effect of student activity on their epistemic emotions is also visible in the results of Study III. Almost half of the students in the data were notably bored throughout the learning task, and many also gave relatively short and vague an- swers in the POE items. This implies that these students were not actively engaged in the learning activity, but were only passively executing the given task. It has also been suggested in previous research that passive participation in learning ac- tivities induces boredom (Minkley et al., 2017). Thus, it could be argued that high cognitive demand and the student’s active role induce negative, activating epis- temic emotions. In turn, low cognitive demand and a passive role may induce boredom.
4.1.2 Associations between students’ epistemic emotions and sci- ence learning
Studies II and III addressed the question concerning the relationship between stu- dents’ epistemic emotions and science learning. A pre-posttest design was adopted in Study II to examine science learning, and in Study III, science learning was conceptualised through scientific sensemaking and examined within the POE ac- tivity.
In Study II, as hypothesized, positive epistemic emotions (curiosity and ex- citement) were found to correlate positively with pretest and posttest performance, and negative epistemic emotions (frustration and boredom) to correlate negatively with pretest and posttest performance. This finding is consistent with results re- ported in previous literature (Ainley & Ainley, 2011; Bosch & D’Mello, 2017;
Gruber et al., 2014; Muis et al., 2015; Pekrun et al., 2014), and thus further con- firms the association between epistemic emotions and performance. However, surprise was not found to relate to performance, even though the results of some previous, experimental studies imply an existing relation between surprise and learning (Chiu et al., 2014; Muis et al., 2018). Furthermore, there was no signifi- cant correlation between test performance and the negatively valenced emotions of confusion and anxiety. However, these emotions may play an interesting role in learning. They had a relatively strong positive correlation with frustration and boredom that relate negatively to performance, but at the same time a clear posi- tive correlation with curiosity that relates positively with learning. Thus, this may indicate that these emotions occur simultaneously in situations, where curiosity is
triggered by new knowledge, but high cognitive demands also cause confusion and anxiety.
Despite the clear correlational relations between epistemic emotions and test performance, the MSEM approach also employed in Study II did not indicate such a strong relationship between situational emotions and learning. Based on MSEM, only boredom had an effect on learning outcome after controlling for prior knowledge. In other words, the more bored students are when studying, the less likely they are to learn. These results are in line with and complement previous findings of a negative relationship between boredom and learning (Pekrun et al., 2014; Tze et al., 2016).
Interestingly, even though curiosity, excitement and frustration correlated strongly with both pretest and posttest performance, they did not relate to change in performance during the study period. However, also the correlation with per- formance could be assumed to indicate something about how they relate to learn- ing. Students with high situational experiences of curiosity and excitement and low experiences of frustration performed better not only in the pretest but also, due to strong autoregression, in the posttest. In other words, at some point these students have either learned more due to a tendency to experience high curiosity and excitement and low frustration, or they have developed a tendency to have these emotional experiences due to their previous performance.
The results of Study III are in line with those of Study II, indicating that levels of epistemic emotions vary significantly especially in relation to students’ ability to make relevant observations. As revealed in a cross-tabulation, students whose emotional trajectories indicate high levels of curiosity and increasing levels of surprise and confusion are better able to make sense of the phenomenon than stu- dents with high levels of boredom. This implies that boredom may have an inhib- iting effect on making observations and developing understanding, as has been suggested in previous research (Muis et al., 2015; Pekrun et al., 2014; Tze et al., 2016). Correspondingly, curiosity, confusion and surprise may foster these cogni- tive processes. As Muis and colleagues (2015) suggest, different learning strate- gies related to these emotions may explain this implication. Activating epistemic emotions such as surprise, curiosity and confusion typically relate to deep-pro- cessing learning strategies. Moreover, curiosity in particular plays a role in critical thinking, related to the application of previous knowledge to new situations and the evaluation of information (Muis et al., 2015). These deep-processing learning strategies also play a central role in the making of scientific observations. In turn, as a negative deactivating emotion, boredom is related to the impairment of any learning strategies, which in the case of Study III appears as the inability or un- willingness to make scientifically relevant observations. Overall, the findings cor- roborate the assumption that emotions influence scientific sensemaking and rea- soning (Fischer et al., 2014; Theobald & Brod, 2021; Wickman et al., 2021).
Overall, the role of confusion in learning is noteworthy. Even if it is a negative emotion based on its valence, in terms of learning it could be considered positive.
It was found in Study II that confusion did not relate significantly to performance or learning, but it correlated positively with curiosity. Furthermore, the regression coefficient from pretest performance to situational confusion in MSEM was neg- ative (and statistically significant), but the regression coefficient from situational confusion to posttest performance was positive (but statistically non-significant, p = 0.07), underlining the complexity of the interplay between learning and situa- tional confusion (D’Mello et al., 2014). As indicated in the findings from Study III, confusion often appears together with the positive emotions of curiosity and surprise, and is thus experienced by students who perform better in scientific sensemaking. These findings are in line with the results of previous research sug- gesting a positive relationship between confusion and learning (D’Mello et al., 2014; Muis et al., 2015; Watkins et al., 2018), but contradict other research find- ings indicating that confusion detracts from learning (Jacobson et al., 2021;
Schneider et al., 2016). Indeed, the present results indicate that situational confu- sion should also be considered a learning enhancer rather than only a detractor.
However, further research should be carried out to establish the role of confusion in science learning.